Moody's Corporation (MCO) Earnings Call Transcript

Good day, ladies and gentlemen, and welcome to the Moody's Data and IT Strategy Conference Call. [Operator Instructions] It is now my pleasure to turn the floor over to your host, George Tong. Sir, the floor is yours.

Thank you, and thank you all for joining us to discuss Moody's data and IT strategy. I'm George Tong, and I cover business and info services at Goldman Sachs. I'm really pleased to be joined by Mona Breed, Chief Information Officer at Moody's. Before we begin, we are required to make certain disclosures in public appearances about Goldman Sachs' relationships with companies that we discuss. The disclosures relate to investment banking relationships, compensation received or 1% or more ownership. These disclosures are available in our most recent reports available on our firm's portals. Disclosures and updates to those disclosures are also available by ticker on the firm's public website at gs.com/research/hedge. Also the views stated by non-Goldman Sachs personnel do not necessarily reflect those of Goldman Sachs. Participants who would like to ask a question can submit it via the webcast or e-mail it to george.tong@gs.com. We'll get to those questions in the second half of the call. Mona, thank you for joining us today.

Thank you very much for having me. It's a pleasure to be here.

Great. So Mona, you serve as the Chief Information Officer for Moody's. Can you describe the responsibilities of your role and your oversight within the company?

I look after core data and technology functions that are shared across the firm, so by -- both the ratings agency as well as our analytics business. Those include cyber, infrastructure, enterprise data and architecture, all of our corporate applications, sourcing and procurement. And more recently, my team was tasked with building out our integration management office, which actually covers the integration activities of all the acquisitions for Moody's.

Got it. So to start, how would you describe Moody's philosophy in overarching strategy around information, data and technology?

I think, first, I'd say, data and technology are really at the heart of our business strategy. And the way we think about it is that we use a whole range of technologies to power integrated data, analytics and workflow solutions that help our customers make really informed decisions about the risk ecosystem in which they operate. So underpinning that are 4 things that we think about. The first is customer centricity. We're focused on innovating and developing new enhanced products that meet the customers' evolving needs, and we want to do that with speed to market. The second thing is we are very deliberate, innovation with intent. So technology is, for me, very exciting and interesting, and there are a lot of different technologies that could be applied to a whole host of business challenges. When we say invest with intent, we really are focused in terms of where we're innovating and where we're putting capital to take things to the next level from an innovation standpoint. Will it really address a particular business challenge? And can we test that out? The third thing that we think about is security, reliability, performance and scalability. If you think about Moody's, trust and reliability are core tenets of our customer value proposition. So we think it's really important to invest in multi-zone, serverless and secure cloud infrastructure. The last thing, and perhaps one of the most important, is culture. We've made a significant investment in creating a technology culture of innovation, collaboration and really bringing people together to solve complex business problems across the firm. This has allowed us to attract, retain and grow some really excellent technology talent.

Got it. Now you mentioned that data and analytics sit at the heart of what Moody's does. And certainly, Moody's makes use of a significant amount of data in both its Ratings and Analytics businesses. What technologies do you use to help with data ingestion?

To set the stage, we have about -- just over 26 petabytes of storage. So that's a pretty vast scope of data sets. We leverage a variety of tools, AWS EMR and Glue are 2 that I'd mentioned because they're used across a number of product suites. We also use Airflow, particularly Astronomer Airflow, it's beginning to pick up speed in a few of our domains. And a recent addition to our data lineup is Snowflake. Primarily, we use that right now for internal provisioning, aggregation and analytics. We like Snowflake because it's single-stop shopping in terms of the data lake, the tooling and the analytics, and it's cloud agnostic. In the cyber space, we've adopted Cribl [ to help us ] with ingestion, particularly with respect to log data, making it a little bit faster and easier to focus on meaningful data sets as opposed to some of the noise that can be captured in that logging.

Got it. And as you think of all that data, certainly, to a large degree, the value-add of Moody's stems from its ability to derive predictive insights from this data that you're managing and storing. What roles do AI and machine learning play in this endeavor in each of the segments? And perhaps you can talk about some of the strategies in MIS and some of the strategies in MA.

Sure. AI and ML are definitely a critical part of the technology strategy across the firm. While we do offer incredible insights, content and data sets, we are equally focused on expanding our top caliber technology solutions and -- that maximize the impact, reach and value of that underlying integrated risk assessment data. So I'll give you one example. In 2018, we launched DAISY, which is a Stevie innovation award-winning solution for our MA customer service. DAISY detects the language, predicts the product and category, and routes customer service inquiries to the right client service support team. It also provides suggestions on answers based on our data set, so as it's learning. And you may be familiar with QUIQspread, which automates the spreading of financial statements. It recently won Best Financial Services AI Solution at the Artificial Intelligence Award. In our real estate practice, we use satellite images and machine learning to monitor new construction activity. This allows us to know when the new construction is occurring, and it replaced a fairly significant manual process and alerts us to movement so we can monitor until that new construction is complete. Finally, in Moody's ESG, we use AI and ML to extract keywords from documents to actually see the metrics calculation. And so we're building an annotation tool as well there that allows real-time modification and re-execution of those ESG models.

Got it. Now capabilities around AI and machine learning can be built up internally or to be acquired. What are some recent acquisitions you've completed that have strengthened Moody's capabilities here?

Excellent. I'm going to start actually with our internal capabilities, and I'll move into those that we've acquired. Back in 2016, we created the Accelerator unit, which actually delivered QUIQspread that I mentioned previously. In this model, we brought together squads of people from across the business and technology to adopt a more agile approach. It was a true incubator, complete with the ability to test ideas, fail fast, deploy new customers using sophisticated DevOps pipelines and constructs. So -- and that's still in full force and a pretty impressive internal unit. If I move to the acquisition side of it, I can mention that from the TCFD status report, they actually use the AI pipeline from ML fabric, which is a custom-designed, modular, reusable cloud platform that we developed at Moody's to operationalize deep learning and machine learning models. So it really allows users to deploy and reuse AI models and AI workflows at scale. So they leveraged the ML fabric models-as-a-service platform to scale the processing of more than 2,000 documents, financial filings and other documents for the selected population of companies to produce those results. And I'll give you one other example. More recently, in terms of AI and ML from another acquisition, if you think about the Acquire Media acquisition. So we take news feeds from Acquire Media to update somewhere between 7,000 and 10,000 profiles a day in the RDC GRID database. Within this KYC solution, AI and ML really help us provide a curated risk-relevant profile by drawing on the deepest risk data without the noise of false positive. So one customer even highlighted the fact that RDC GRID reduced false positive in their use case by about 70%. Another customer noted the power of 5x faster screening with continuous monitoring and the value that adds to their business.

Got it. Now what particular technologies would you be interested in further building out through M&A at Moody's? What's cutting edge out there that you hope to bring into the company?

Yes. And as I discuss and think about our M&A strategy, and it's really around adjacencies, both from the data perspective as well as from the technology perspective, so how can we create interoperable technology solutions that solve real business challenges, being able to maximize those M&A adjacencies by combining data, services to help customers navigate risk. So as we're approaching M&A, I don't think about it in terms of a particular technology or a particular data set. It's really looking at that whole integrated risk assessment, operating system and where are there areas that we're right to sort of expand. So if you look at the most recent announcement of PassFort and kompany M&A for the KYC practice, both of these are really -- both of these organizations are really innovators in the compliance and regulatory space. And their tech solutions will upgrade and accelerate our customers' onboarding and monitoring processes. So we're really excited about that. If you think about PassFort, it's a SaaS-based workflow platform that -- for identity verification, customer onboarding and risk analysis. And it's really expected to help us at Moody's create more holistic workflow solutions, and customers can then incorporate Moody's data, including credit, cyber, ESG and climate analytics directly into their proprietary processes. So that's very exciting to us. Similarly, the company's ensuing acquisition will bring technology and APIs that enable Moody's customers to complete shareholder analysis and entity verification in real time. Finally, I'll talk about Cortera. And when we think about interoperability, and it's back to the comment I made about the adjacencies, is we are making acquisitions, wanting to bring together data and technology sets. And Cortera is example where they were able to introduce the COVID-19 Economic Impact Tracker. They actually did that by leveraging Economy.com data capabilities. So it's really that interoperability of these adjacencies, if you will, to create new solutions to help our customers manage those risks.

Right. Now can you talk a little bit about how technology helps with innovation at Moody's? What are some examples of how your tech stack enables more efficient new product innovations?

Yes. So first, technology really starts with a solid core. And I will share that over the past 3 years, we've really, really invested and focused on modernizing the foundation and creating what we refer to as a shared approach to our architectural choices or more federated approach to how we build certain things in the product engineering space. So we did this deliberately to promote building state-of-the-art shared component. So if you think about things like single sign-on UI, UX, CI, CD, it was why do we have product teams potentially building solutions in those classes across each product engineering team, why not think about building a top-notch, world-class offering that multiple product teams can use. So they are actually then, those product engineering teams, very focused on the innovation and collaboration required to take those products to the next level. So we've really focused on how can we leverage certain capabilities across the product suite, and that has really allowed us to reinvest a lot of the product engineering's time into product development as opposed to rebuilding components that could be built once and used many times.

Right. Now where is Moody's with respect to its cloud migration? How much of the company is now based in the cloud? And what are next steps as you think about your cloud strategy?

So for the Moody's Analytics practice, we began our cloud journey, and I guess, for shared services as well back in 2015. And today, we have about -- a little over 90% of the MI estates in the cloud and a large percentage of our shared services. Application SaaS solutions are also in the cloud. We -- in the MIS part of the business, we are starting that public cloud journey. We're largely on-prem with some work in more of a private cloud setup. And along the way, we learned a lot of critical lessons about sort of workload refactoring. When we originally went to the cloud, it was really in an effort to reduce our data center footprint. And so most of the migrations were done in more of a lift and shift fashion. And we really discovered after doing that that we had some work to do around some refactoring so that those solutions would be both cost effective and performant in the cloud. So those lessons were learned, and they will be applied as our MIS business begins its cloud journey. As we start to think about the future, I guess, today, we're very focused on serverless patterns. It makes it easier to migrate from one cloud provider to another. It also helps with cost efficiencies. So serverless is top of mind right now. As we think in out years, it's really what are the use cases that would drive the implementation of a true multicloud solution. And that may apply to some of our products. It may not apply to all of them. But that's -- those are the things we're starting to think about. The good news is our migration to serverless will actually make that migration to multicloud more feasible and seamless for those use cases as they are determined.